(a). Field of the Invention
The present invention relates in general to auto-negotiation in a network, and more particularly to a repeater and associated method for performing auto-negotiation between two node devices in the network.
(b). Description of the Prior Arts
A host computer usually connects to a local area network (LAN) via a network interface card (NIC). For an Ethernet network, there are several available technologies, such as 10Base-TX, 100Base-TX, and 100Base-FX. The PHY of the NIC capable of these technologies supports detection of capabilities of a connected network node and determination of a highest performance mode of operation between them. This function is called auto-negotiation and defined in IEEE 802.3u standard.
A mode of operation of a NIC comprises data rate (e.g. 10 Mbps or 100 Mbps), Half/Full Duplex, and flow control. Full Duplex mode means that the NIC can perform data reception and transmission on a network simultaneously, while Half Duplex mode means that the NIC can only perform one of data reception and transmission in the same period of time. Flow control refers to the capability to control the traffic on the network.
The auto-negotiation function allows two connected node devices to automatically detect the PHYs each other, thereby knowing modes of operation supported by the opposite side and determining the highest performance common mode between them. A node device, such as a NIC or a switch, can configure an auto-negotiation advertisement register (ANAR) in the PHY through a serious management interface of media independent interface (MII), wherein the ANAR is used to record related information to be advertised to remote devices from the host device such that the remote devices can know the capabilities of the host device when performing auto-negotiation.
FIG. 1 is a diagram illustrating protocol implementation for auto-negotiation in an Ethernet network. As shown in the right of FIG. 1, the Open System Interconnection (OSI) reference model, which is the most commonly used layering model for network, comprises a Physical layer 1, a Link layer 2 and other layers. In the left of FIG. 1, the Physical layer 1 and the Link layer 2 correspond to respective layers used in CSMA/CD of IEEE 802.3, and auto-negotiation is implemented in the Physical layer 1 of the OSI model. A PHY device 10 is located between a media independent interface 11 and a media dependent interface (MDI) 12. The MDI 12 connects to other network devices via a physical transmission line 13. The Link layer 2 comprises a logic link control (LLC) layer 20 and a media access control (MAC) layer 21.
The PHY 10 comprises a physical coding sublayer (PCS) 101, a physical medium attachment (PMA) layer 102, a physical medium dependent (PMD) layer 103, and an auto-negotiation layer 104 which detects other PHYs when performing auto-negotiation. The details of auto-negotiation are further described in IEEE 802.3 standard.
In an Ethernet network, the maximal allowable distance between two nodes is only 100 meters. As shown in FIGS. 2a and 2b, the distance between NICs 30, 31 and between switches 32,33 is limited to within 100 meters. Therefore, a repeater (or several repeaters) is needed to maintain a network connection between two distant nodes, e.g. a host computer and a remote server.
However, a conventional repeater can only strengthen signals and is incapable of auto-negotiation. If a node connected to the conventional repeater changes its data rate, e.g. from 10Base-TX to 100Base-TX, the data rate would be still limited to 10 Mbps since the conventional repeater cannot perform auto-negotiation to adjust its own data rate. At this time, a manual adjustment would be needed to change the data rate of the conventional repeater to 100 Mbps if the conventional repeater also supports a 100 Mbps data rate.